From: "Nikolaos D. Bougalis" <nikb@webmaster.com>
To: <netdev@vger.kernel.org>
Subject: RFC: Established connections hash function
Date: Thu, 22 Mar 2007 08:39:04 -0700 [thread overview]
Message-ID: <10189ABA61CF4D5AB3881F96C9CACE87@XEON> (raw)
Hello,
I have noticed that the hash function that the kernel uses for
established TCP/IP connections is rather simplistic, specifically:
h = (local address ^ local_port) ^ (remote_address ^ remote_port);
h ^= h >> 16;
h ^= h >> 8;
Now, simple is great, but this has a number of issues, not the least of
which is that an attacker can very easily cause collisions and force
extremely long chain lengths, a situation that becomes worse the more
distinct IP addresses and listening ports a box has.
Consider, for example, a box that has 20 ports open and 4 consecutive IP
addresses. An attacker that has an entire class C available can create
24,576 connections that hash to the same value, resulting in a ridiculously
overlong chain. With servers that do virtual hosting and have dozens of IPs,
the situation can become much worse very fast.
This particular hash seems to be the odd-man out, since most other
network related hashes in the kernel seem to be Jenkins-based, and some use
tagged hashing to defeat algorithmic complexity attacks. For example, the
route hash uses this:
static unsigned int rt_hash_rnd;
static unsigned int rt_hash_code(u32 daddr, u32 saddr)
{
return (jhash_2words(daddr, saddr, rt_hash_rnd)
& rt_hash_mask);
}
With this in mind, I propose the following replacement for inet_ehashfn,
which defeats algorithmic complexity attacks and achieves excellent
distribution:
unsigned int inet_ehashfn(const __be32 laddr, const __u16 lport,
const __be32 faddr, const __be16 fport)
{
return jhash_3words((__force __u32)faddr, (__force __u32)laddr,
(((__force __u32)fport) << 16) + lport,
inet_ehash_rnd);
}
where inet_ehash_rnd is initialized once in tcp_init to a random 32-bit
value.
I will be more than happy to provide a patch for this, but I figured I
would solicit some input first.
Nik B.
next reply other threads:[~2007-03-22 15:44 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-03-22 15:39 Nikolaos D. Bougalis [this message]
2007-03-22 15:52 ` RFC: Established connections hash function Evgeniy Polyakov
2007-03-22 17:32 ` Nikolaos D. Bougalis
2007-03-22 18:21 ` Evgeniy Polyakov
2007-03-22 19:44 ` Nikolaos D. Bougalis
2007-03-22 19:56 ` Evgeniy Polyakov
2007-03-22 20:53 ` Nikolaos D. Bougalis
2007-03-23 7:52 ` Evgeniy Polyakov
2007-03-22 20:58 ` David Miller
2007-03-22 22:03 ` Eric Dumazet
2007-03-23 7:11 ` David Miller
2007-03-23 8:00 ` Eric Dumazet
2007-03-23 18:46 ` David Miller
2007-03-23 8:07 ` Evgeniy Polyakov
2007-03-23 8:17 ` Eric Dumazet
2007-03-23 8:33 ` Evgeniy Polyakov
2007-03-23 9:10 ` Evgeniy Polyakov
2007-03-23 11:58 ` XOR hash beauty solved [Was: RFC: Established connections hash function] Evgeniy Polyakov
2007-03-23 12:51 ` Nikolaos D. Bougalis
2007-03-23 12:45 ` RFC: Established connections hash function Nikolaos D. Bougalis
2007-03-27 14:11 ` Andi Kleen
2007-03-28 5:01 ` Nikolaos D. Bougalis
2007-03-28 6:29 ` David Miller
2007-03-28 9:29 ` Andi Kleen
2007-03-28 10:45 ` Evgeniy Polyakov
2007-03-28 14:14 ` Andi Kleen
2007-03-28 13:50 ` Eric Dumazet
2007-03-28 14:52 ` Andi Kleen
2007-03-29 9:18 ` Evgeniy Polyakov
2007-03-28 14:17 ` RFC: Established connections hash function II Andi Kleen
2007-03-28 19:04 ` RFC: Established connections hash function David Miller
2007-03-28 20:12 ` Andi Kleen
-- strict thread matches above, loose matches on Subject: below --
2007-03-24 12:26 linux
2007-03-24 13:29 ` Evgeniy Polyakov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=10189ABA61CF4D5AB3881F96C9CACE87@XEON \
--to=nikb@webmaster.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox